US government health report "riddled" with AI errors
A White House report on the decline in Americans’ life expectancy has been widely criticised for containing numerous citation errors, fabricated studies and signs of AI-generated content, undermining its scientific credibility. What happened Led by Health and Human Services Secretary Robert F. Kennedy Jr., t he White House’s “Make America Healthy Again” (MAHA) report was filled with citation errors, duplicated references, incorrect author attributions, and references to studies that do not exist, according to investigations by NOTUS , The Washington Post and other news publications. At least seven cited studies were entirely fabricated, and some URLs included markers such as “oaicite,” which is commonly added by OpenAI’s AI systems. The report also misinterpreted or misrepresented real studies, further eroding trust in its findings. The errors are seen to have significant potential harms: policymakers, healthcare professionals, and the public may be misled by inaccurate or nonexistent data, thereby resulting in flawed policy decisions or public health initiatives. Why it happened Experts and analysts have identified clear evidence that artificial intelligence - likely tools like OpenAI’s ChatGPT - was used to generate or supplement the report’s references and citations. AI systems are known to “hallucinate” or invent sources, blend real and fake information, and make formatting errors, which explains many of the report’s issues. The White House and US Department of Health and Human Services attributed the errors to “formatting issues,” but independent experts and journalists have pointed to AI as the primary cause. The rapid compilation of the report - completed in just over three months - and a lack of thorough review before publication likely contributed to the oversight. What it means The incident raises concerns about the government’s commitment to evidence-based health policy and the transparency of its processes, as well as the risks of over-reliance on AI in scientific and government reporting. Society may become more skeptical of government health reports and AI-generated information, potentially reducing trust in official sources of public health guidance. System 🤖 ChatGPT Developer: OpenAI Country: USA Sector: H ealth Purpose: Generate academic citations Technology: Generative AI; Machine learning Issue: Accuracy/reliability; Mis/disinformation; Transparency Resource s 📃 White House. The MAHA Report
- Date it happened
- 2025-01-01
- Organisation involved
- US Department of Health and Human Services
- Product, system or model
- ChatGPT
This incident was imported from AIAAIC and is used under CC BY-SA 4.0. Our additions to it — the structured fields, the translation, the checks against other reports — are published under the same licence.
This is a record of what was reported, not a finding that anyone broke the law. If it names your organisation and you believe it is wrong, the corrections process is free and open to everyone.